lectures.alex.balgavy.eu

Lecture notes from university.
git clone git://git.alex.balgavy.eu/lectures.alex.balgavy.eu.git
Log | Files | Refs | Submodules

Control systems.html (3113B)


      1 <?xml version="1.0" encoding="UTF-8"?>
      2 <!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">
      3 <html><head><link rel="stylesheet" href="sitewide.css" type="text/css"><meta http-equiv="Content-Type" content="text/html; charset=UTF-8"/><meta name="exporter-version" content="Evernote Mac 7.5.2 (457164)"/><meta name="altitude" content="2.2164626121521"/><meta name="author" content="Alex Balgavy"/><meta name="created" content="2017-11-06 12:38:55 +0000"/><meta name="latitude" content="52.33300980610338"/><meta name="longitude" content="4.865525386192134"/><meta name="source" content="desktop.mac"/><meta name="updated" content="2017-12-19 19:46:43 +0000"/><title>Control systems</title></head><body><div>Sensors: receive signals (microphones, cameras…)</div><div>Actuators: actually do stuff in the 'real world’ (LEDs, motors, speakers, displays, lamps…)</div><div>Controller: “the brain”, the intelligent unit. provide hardware &amp; software that makes system autonomous by using sensor input etc.</div><div><br/></div><div>any pervasive computing system executes sense-control-act sequence in a loop</div><div><br/></div><div>Control:</div><div><ul><li>deliberative: think hard, act later</li><ul><li>planning — look ahead at outcomes of possible actions</li><li>searching — looking for sequence of actions leading to desired goal</li><li>use internal representation of the environment — a map, for example</li><li>for decisions, use for example shortest path from one node to another in a map</li><ul><li>uses Dijkstra’s or A* algorithm</li><li>GPS nav systems use this</li></ul></ul><li>reactive: don’t think, react!</li><ul><li>e.g. smart curtains, thermostat, obstacle handling, landmark navigation</li><li>don’t use internal representation, just direct mapping between sensors and effectors</li><li>rules:</li><ul><li>if dark outside, then close the curtains</li></ul><li>control type</li><ul><li>open-loop control — input signal to controller, actuator, output controlled variable</li><ul><li>examples: microwave, automatic lights, automatic water faucets</li></ul><li>closed-loop control — get feedback, check if everything was executed right</li><ul><li>uses a comparator that gets feedback from output of actuator</li><li>comparator outputs error to controller, which then tries to minimise error</li><li>example: heater</li></ul></ul><li>obstacle handling</li><ul><li>simple — contact (touch sensor)</li><li>better — proximity, but don’t know distance (whiskers)</li><li>best — ranging</li><ul><li>sonar, with reflected sound waves. echolocation (distance = speed × time)</li><li>LIDAR, using a laser swept across FOV</li></ul></ul><li>landmark navigation</li><ul><li>follow a line, a wall</li><li>feedback control — turn always same angle, turn proportionally (P), proportional derivative (PD, rate of change), or proportional derivative and integral (PID, rate of change and time)</li></ul></ul><li>others: hybrid, behaviour-based</li></ul><div><br/></div></div><div><br/></div></body></html>